South Arabia’s prehistoric monument landscape shows social resilience to climate change

PLOS One | https://doi.org/10.1371/journal.pone.0323544 May 28, 2025 5 / 24 and former caravans [30,61], something we documented during 22 years of fieldwork in Hadramawt and Dhofar. These routes may also have higher monument frequencies than off-route, difficult- to-access locales, which we examined by adding our foot tracks and our meanders to reach targets generated by a semi-automated detection of monuments [62]. Where the survey team’s passage was rapid (i.e., major asphalt roads), we eliminated tracks from subsequent analysis; but we documented each monument we observed by a simple inventory list when time, resource and logistics prevented more thorough investigation. Survey over multiple years and projects nevertheless consistently collected all monument data within defined viewsheds (sampled areas). Between 2009–2018, we surveyed 428 monuments. Opportunistic, judgmental sampling drawing on the expertise of crews with decades of collective field experience was complemented by collaboration with faculty and graduate students in engineering and statistics to develop algorithms that detected monuments in satellite imagery [62,63]. Our logged tracks became transects in which we recorded the presence and absence of monuments. This network of monuments and tracks served as the basis for viewshed analysis, a computation that maps the areas of the surrounding landscape visible from a given point or points based on whether those areas lie above the local landscape horizon [64,65]. We used an ASTER

Global Digital Elevation Model (GDEM) (30 m resolution) and observer height set to 1.6 m to identify all areas visible from the locations we visited. Based on extensive field experience in which we measured maximum distance to sighted monu- ments, we set an 800 m viewshed boundary to define a full-survey coverage area. In conformity with ethical standards, we conducted all field research under (un-numbered) permits issued annually through 5-year agreements signed between Ministry of Heritage and Tourism, Sultanate of Oman and Joy McCorriston (2007–2013, 2017–2021). Field documentation methods. We documented monument attributes, including location, other visible monuments, materials, preservation, cultural type, and measurements to calculate stone volume and largest stones as proxies for large group sizes and simultaneous labor input [66]. We differentiate between episodic monuments (built in one episode) and accretive monuments, which can be assembled in repeated visits to a particular location. Our terminology refers to the initial construction of monuments according to a structural-cognitive template. We recognize, as both stratigraphic examples and in theoretical concept, that monuments were re-visited, altered, additions made after substantial lapses in time, sometimes partially or fully dismantled, and appropriated—such is the role of monuments in both shaping and being shaped by social groups as literal touchstones for institutional thought [41,67]. For example, through radiocarbon

determinations, we documented instances of animal offerings placed into tombs opened and re-sealed a thousand years after a dated burial. This use is unrelated to initial construction, which was the target of our documentation through chronology and measurements. Using high-precision GPS and tape measures, we established geographic coordinates on the center of a monument, its perimeter or length, and took compass bearings on orientation [59,62]. For all monuments, we measured the length of the largest stone, registered as a categorical variable greater or less than 0.25, 0.50, 0.75, 1.0 m. These ranges index labor group sizes from single to many people. For most small-scale stone monuments, the largest measured stone could be shifted by the two to four adults in one herding family, ethnographically averaging five to seven adults and children [30,68]. Our experience with excavation crews dismantling collapsed monuments in both Yemen and Oman, shows that to shift limestone blocks greater than 0.75 m length minimally requires seven or more adults, particularly as tools such as iron levers were unavailable in prehistory. Transport of such stones even a few meters implies the cooperation of a substantial group. We photographed each monument from multiple angles, which allowed us systematically to generate standardized estimates of stone volume from geometric shape and dimensions. We calculated the volume of stone using volumetric calculations for six geometric shapes, which closely match the monument shapes. Platforms can be cylinders

as well as trapezoidal shape; SCAB monuments are cone-shaped; HCT are cylindrical x 0.67 to account for a hollow chamber; triliths are the shape of a triangular prism x 0.5 to account for empty space; and boat graves are straight-sided lozenges.